Objective:To evaluate the correlation between late spermatid score and spermcount in men with normal semen analysis and oligozoospermiaundergoing ICSI and setting the average threshold of spermatids neededfor normal sperm count.Materials and methods:This retrospective study involves reevaluation of testicular biopsy slidesof 42 subjects, 18 of them were oligozoospermic and 24 subjects withnormal semen analysis who underwent wide bore needle testicular biopsyat time of ICSI due to failure of semen collection.Results:We found a significant correlation between sperm count, late spermatidscore and Johnsen score in normozoospermic subjects. But Inoligozoospermic subjects, we found a correlation between sperm countand late spermatid score but not with Johnsen score. We found inoligozoospermic group with more than 10.000.000 sperm/ml a significantcorrelation between sperm count, late spermatid score and Johnsen score.But in the group with less than 10.000.000 sperm/ml, there is nosignificant correlation between sperm count, late spermatid score andJohnsen score. The threshold of late spermatid score in predicting morethan 20 million sperm/ml was 20.75. Coefficients of correlation justfailed to reach the border of statistical significance betweennormozoospermic subjects and oligozoospermic patients regardingbasement membrane thickness, but there was significant differenceregarding seminiferous tubule diameter. There were no relevanthistopathological changes regarding edema and fibrosis in both groups.Conclusion:The quantitative analysis of testicle biopsy is a simple technique.Accordingly, late spermatid score can correlates closely with spermcount. It can be of great value in evaluation and follow-up of patientswith oligozoospermia or obstruction azoospermia.