We have previously shown that rats given medial prefrontal lesions on postnatal Day 10 show a performance deficit on spatial delayed alternation (2‐s delay) when tested between Days 19 and 23, but show recovery of function by Days 27–33 (Freeman & Stanton, 1992). The present study attempted to (a) examine the effects of extending the delay interval from 2 s to 15 or 45 s on the ontogeny of spatial delayed alternation; and (b) determine whether these longer delays would alter the ontogenetic profile of recovery from medial prefrontal damage. Animals were given medial prefrontal lesions (mPFC) or sham surgery on Day 10. They received acquisition training on delayed alternation with a 2‐s delay on Day 21 or Day 28, followed by “probe” tests at delays of 15 and 45 s on the following day (Day 22 or Day 29). On Days 21–22 there was an overall effect of lesion at the 2‐, 15‐, and 45‐s delay intervals indicating and early deficit. However, on Days 28–29 there was no effect of lesion at the end of acquisition or at the 15‐ or 45‐s delay intervals. Recovery of memory function by mPFC animals on this task appears to occur during the weanling period regardless of delay interval. © 1995 Johnv Wiley & Sons, Inc. Copyright © 1995 John Wiley & Sons, Inc.