Temporal lobe epilepsy is the most common form of refractory epilepsy in adults, accounting for approximately 60% of cases, and is one of the most difficult epilepsies to cure medically. The surgical treatment of refractory epilepsy is the simplest and most classic in temporal lobe epilepsy, and the postoperative seizure control is mostly good. With the further development of preoperative examination techniques such as neuroimaging and brain electrophysiology, the outcome of epilepsy surgery has been significantly improved again. However, the outcome of epilepsy involves many preoperative factors, and there are few systematic summary reports on what factors may predict a good outcome. We summarized the results of 143 patients with temporal lobe epilepsy who underwent surgery as follows: (1) Neuroimaging factors MRI is able to detect subtle changes in cortical structures and changes in tissue signal intensity, and is highly sensitive and specific for identifying various types of pathological foci in symptomatic epilepsy. In temporal lobe epilepsy, hippocampal sclerosis is a common etiology, which is demonstrated in MRI: (i) hippocampal atrophy coronally shows loss of normal oval shape and is thin and flat, unilateral hippocampal atrophy, which is smaller than the contralateral side by more than 30%, showing bilateral hippocampal asymmetry; (ii) T2-weighted image is high signal, especially on Flair sequence; (iii) temporal horn enlargement; and (iv) anterior temporal lobe atrophy. It should be suggested that the Flair sequence suppresses the cerebrospinal fluid signal, which can avoid the interference of partial volume effect and mobility artifacts produced by cerebrospinal fluid, while highlighting the long T2 component, forming a high contrast between the lesion and the background, allowing the small lesions located in the paraventricular and gray matter areas, especially in the hippocampus (medial temporal horn of the lateral ventricle) to be more clearly displayed. In 27 cases, lesions located in the paraventricular and gray matter regions, which were not shown on conventional T2 images, were clearly shown on the Flair sequence. In addition, the sensitivity and specificity for low-grade malignancies and cortical dysplasia were also high. The satisfactory postoperative outcome rates of 102 cases with and 41 cases without structural abnormalities suggested by MRI were 88.2% and 75.6%, respectively, and the comparison between the two was statistically significant, suggesting that the postoperative outcome of the former was better than that of the latter. Among them, the postoperative outcome satisfaction rate of 41 cases of unilateral hippocampal sclerosis reached 95.2%, however, the bilateral sclerosis was only 78.2%. The former can be removed more completely by surgery, while the latter can actually be solved only on one side in one operation, which actually belongs to “palliative surgery”. The highest rate of satisfactory postoperative outcome was found in unilateral cases, which may be a reliable prognostic factor. The good postoperative outcome in this group of patients with complex partial seizures as the main manifestation and with onset before 5 weeks of age may be due to the fact that the former is characteristic of medial temporal lobe epilepsy, and the early onset of seizures is also a high risk factor for hippocampal sclerosis (volume reduction). All of them reached a satisfactory level, suggesting that the prognosis of surgery for benign tumors secondary to epilepsy, which are commonly seen in young patients, is good. Low-grade malignant tumors are associated with a localized epileptic origin and are limited in extent, so the outcome after surgical resection is significant. The reason for this is that softening foci are often secondary to traumatic brain injury or ischemic damage and often coexist with other structural abnormalities on the contralateral side, and treatment of the epileptic foci around the softening foci can cause further deterioration of already low neurological function, making it difficult to achieve a satisfactory outcome. The abnormalities of cortical dysplasia (e.g., gray matter ectopic) manifest as increased neuronal and white matter clustering, neurogliosis, and perivascular arrangement, mostly in a widespread distribution. This abnormality is located deep in the brain and craniotomy is not possible to reveal and treat. The satisfactory postoperative outcome rate of our five cases of cortical dysplasia was 40.0%, confirming the poor surgical outcome in cases with cortical dysplasia. Theoretically, epilepsy secondary to arteriovenous malformation should be treated after the malformation is treated and the “blood theft” is relieved, and the efficacy should be significant, but the satisfactory rate of the five cases in this group was only 80.0%. However, the number of cases is still small, which is not enough for evaluation. (2) Brain electrophysiological factors Both temporal lobes may become epileptic areas, and the location of the epileptogenic focus, especially the main lesion, cannot be directly determined by imaging alone, but it is important to determine the extent of epilepsy for epilectomy. In our group, the excellent postoperative outcome rate was 93.1% in 116 patients with unilateral epileptogenic foci only, especially those with epileptogenic foci and limitations reached 94.6%, which can predict a high satisfactory postoperative outcome rate and confirm that the surgical results should be quite satisfactory if the epileptic region can be accurately localized and completely excised; in addition, there were 51 cases with secondary epileptogenic foci in addition to the main epileptogenic foci, 27 of which were located in the contralateral temporal lobe. The satisfactory postoperative seizure control rate was only 48.1%, far inferior to the former. The reason for this is that epileptic activity in the contralateral temporal lobe is still present and becomes a major factor for postoperative seizure recurrence. In well-selected cases, the rate of complete disappearance of seizures in unilateral hippocampal sclerosis and occupying lesions was significantly higher in this group (92.6%-95.2%). However, the presence of multiple epileptogenic foci in the temporal lobe bilaterally or patients with dual pathology (hippocampal sclerosis and structural lesions outside the hippocampus or even the temporal lobe, with bilateral temporal lobe structural abnormalities) are not absolute contraindications to surgery. In our group, there were 31 such patients, whose epileptic foci were unlikely to be treated all at once, and the surgery was in fact “palliative”. Therefore, it is unlikely that complete postoperative seizure disappearance will be obtained, and only significant remission of seizures can be expected. In our group of 27 patients with multiple bilateral epileptogenic foci, 20 cases were dominant on one side, and craniotomy was performed on that side, and the postoperative outcome of Engel grade II or above was only 55.0%, but grade III or above was 80.0%. Therefore, selecting good cases for surgery can still achieve certain therapeutic purposes. The postoperative satisfaction rate of patients with high seizure frequency (≥20 seizures/month) in our group was low, only 70.9%. The frequency of seizures is actually the success rate of propagation of epileptic activity from the epileptogenic focus to surrounding structures. High seizure frequency implies the presence of multiple epileptogenic foci or channels that can easily propagate to the periphery, often accompanied by bilateral presence of epileptogenic foci, so the surgical outcome is reduced. Of the 102 patients with pathological foci detected by MRI, the satisfactory surgical outcome was significantly better in 94 cases (92.6%) in which the pathological foci were located on the same side as the main epileptic foci derived from dipole analysis than in those in which they were located on different sides (37.5%). This suggests that the ability of structural and electrophysiological localization to be consistent can predict excellent postoperative results. (3) Surgical factors We mainly adopted anterior temporal lobectomy in the early stage of epilepsy surgery, and later found that this type of surgery was too extensive for simple medial temporal lobe epilepsy patients, which increased the incidence of complications such as aphasia. Taking advantage of the high resolution of MRI and dipole analysis for pathological and epileptogenic foci, respectively, we gradually adopted selective hippocampal amygdala resection, simple resection of the lesion or additional hippocampal resection for those with lesions in the posterior temporal lobe. In our group, the satisfaction rates of 86.5% and 81.5% for anterior temporal lobectomy (111 cases) and selective amygdala resection (27 cases), respectively, were not statistically different. However, this does not mean that the choice of procedure is arbitrary, because the degree of epileptic foci removed during epilepsy surgery determines the postoperative outcome, and the limited field in the latter procedure may leave a certain amount of epileptic foci in other areas unexposed and untreated, and the residual epileptogenic foci may be the source of postoperative recurrence. The residual epileptogenic foci may be the source of postoperative recurrence. Although the seizure foci are unilateral, most of the seizures do not improve because of incomplete resection of the posterior hippocampus. It is reasonable to suggest that posterior temporal lobe resection alone or additional hippocampal resection (5 cases) could address both posterior and medial temporal lobes, but the number of cases is not sufficient to make a conclusion. In summary, we can tentatively conclude that although the manifestations of temporal lobe epilepsy are complex and surgical outcomes vary widely, certain preoperative factors may predict a satisfactory postoperative outcome. These include: early onset of seizures, low seizure frequency, no history of persistent status epilepticus seizures, complex partial seizures as the predominant form, MR showing the presence of unilateral structural abnormalities without cortical dysplasia, dipole analysis showing limited epileptogenic foci, and pathological foci consistent with epileptogenic foci.