We have used a sample of 210 UK investment trusts to test performance persistence in different time periods. We bootstrapped on the monthly returns over successive years over the whole period starting from 01/01/1990 to 01/01/2006. The estimated and simulated return was computed by sampling without replacement. The test is performed 210 times over the actual data through a regressing estimated and simulated data. There is evidence of performance persistence over the successive years among estimated and simulated average return of winners and losers computed through the bootstrap test of 210 iterations. There is evidence that publicly available information has been incorporated in the prices of the funds. Most of the observations through two year interval show a positive and significant t-statistic of average estimated return. The average simulated return obtained from the bootstrapped test shows a significant t-statistic. The sample evidence suggests that there is evidence of performance persistence over the successive years among estimated and simulated average return of winners and losers. The test shows that persistence could be explained by long term mean reversion process through replication without replacement.