Two groups of participants attempted eight examples of each of four different problem types formed by combining insight versus non-insight and verbal versus spatial factors. The groups were given different verbalization instructions viz., Silent (N=40) or Direct Concurrent (N=40). There were significant differences between insight and non-insight tasks and between spatial and verbal tasks in terms of solution rates and latencies. Significant interactions between the verbal versus spatial factor and verbalization condition on solution rates and latencies reflected a greater (negative) effect of verbalizing on spatial as against verbal problems. However, no significant interactions of the insight versus non-insight factor with verbalization condition on solution rates or latencies were found. These results favoured the `business as usual' view of insight problem solving as against the `special process' view which predicted larger effects of verbalization for insight problems as against non-insight problems.