Objective To determine the importance of perineural invasion (PNI) on diagnostic biopsy in men enrolled in active surveillance (AS). Patients and Methods Eligibility criteria for AS included clinical stage ≤ T2a and Gleason score ≤6, ≤3 cores positive, maximum single core involvement <50%, and total tumour volume ≤5% on diagnostic biopsy. All men received 12-core confirmation biopsy at ≤6 months. AS 'failure' on confirmatory biopsy was defined as failure to meet one or more eligibility criteria. Risk of AS failure was compared in men with and without PNI. Results For the 165 men comprising the study population, the mean (sd) age was 66.9 (6.5) years and the median (interquartile, IQR) PSA level of men at study entry was 4.4 (3.2-6.0) ng/mL. The median (IQR) follow-up was 5.5 (1.1-9.9) months. In all, 8.5% (14/165 men) had PNI on diagnostic biopsy. Compared with those without PNI, men with PNI tended to have more cores involved with cancer, at a mean (sd) of 2.0 (0.7) vs 1.6 (0.8) cores (P = 0.08) but did not have significantly a greater mean (sd) total tumour length on diagnostic biopsy, at 3.0 (2.1) vs 2.3 (3.6) mm (P = 0.27). Men with PNI on diagnostic biopsy were significantly more likely to meet criteria for disease progression on confirmatory biopsy (57% [8/14] vs 21% [32/151]; P = 0.006). PNI remained a significant predictor for AS failure after adjustment for number of positive cores, maximum percentage core involvement, and total tumour length (odds ratio 4.4, 95% confidence interval 1.4-14.2). Conclusions PNI on diagnostic biopsy is associated with disease progression on confirmatory biopsy. The presence of PNI should factor into appropriate patient selection and counselling in AS. © 2013 BJU International.