The sovereign credit rating is a key determinant of the cost and availability of international financing for an economy. This paper models ratings as a function of expected repayment capacity, derives testable hypotheses, and conducts a statistical analysis based of the ratings awarded by Institutional Investor. The key findings are as follows: 1) Expected rating revisions should be positive for moderately low rated countries and negative for moderately high rated countries. 2) Rating revisions are serially correlated with about one third of a country revision expected to carry-over from one semester to the next. This finding is confirmed with ratings awarded by Moody's and Standard and Poor's. 3) There are regional factors in revisions with about 17 percent of the revision to a regional portfolio expected to carry-over to each country in the region next semester. 4) The serial correlation of revisions is the highest in Emerging and Eastern European countries and the lowest among OPEC members. 5) The 1980s were surprisingly bad years for low- and middle-income country creditworthiness, while the early 1990s were surprisingly good. 6) The East Asian crisis of 1997-98 was much less significant in terms of its effect on global creditworthiness than the debt crisis of the early 1980s.