As expected, it was found that mainly graphic and tabular methods were used, unless the papers in question were purely descriptions of methods or brief case reports. This is probably connected with the long courses of orthopedic conditions, which, with a reasonable investment in time, can only be studied retrospectively. Nevertheless, it became apparent that the group of prospective studies, though numerically smaller, had by and large been performed quite well. Many of them managed without test statistics and yet have considerable information value. The few reports with test statistics had for the most part also been quite well conducted. The most common sources of errors in these were that too many tests were conducted on the same data material without taking the cumulative probability of error into account, e.g., according to Bonferroni-Holm; and the test method used was often not mentioned. So some good statistical studies in orthopedics are certainly being published, albeit gradually. It is planned to conduct a similar investigation on the same years' issues of a German-language journal in another specialty included in the list mentioned at the beginning--German to avoid the bias of translation. Afterwards the two studies can be compared to establish whether an orthopedic journal cannot also be included in the list of the 200. The necessity of impeccable statistics for practice and research is undisputed. The results presented here are intended to encourage orthopedists to attempt prospective studies more frequently than hitherto, and, keeping the test preconditions in mind, also to use correctly described, conclusive statistics.