We conducted a global study of the long-term issuer ratings of nonfinancial firms from Standard and Poor's Ratings Services (S&P) for the period 1998-2003. Specifically, we focused on the solicited versus unsolicited ratings and sample-selection bias in the analysis. Unlike the literature, we adopted an improved method using Wooldridge's instrumental-variable approach to mitigate the concern of specification errors in Heckman's model. We found that the probability of seeking a long-term issuer rating is positively related to the size and profitability of the firm, and negatively related to the growth opportunities and debt levels of the firm. The credit rating is positively related to the sovereign rating, size, and profitability of the issuer, and negatively related to the debt ratio of the issuer. Consistent with the literature, we found sample-selection bias in credit ratings. Our findings suggest that the firms with solicited ratings seem to be more profitable, more liquid, and have lower leverage than the issuers with unsolicited ratings. After controlling for sample-selection bias and some key financial ratios, we found that unsolicited firms, on average, seem to have lower long-term issuer ratings.