This study analyzes the efficiency of internal business transactions by comparing the effect of intra-group sale and purchase transactions on business performance, using listed companies that engaged in transactions between sales-related subsidiaries from 2005 to 2010 as subjects. We classify large and medium business groups based on whether a business group is on the mutual investment restriction list, which the Fair Trade Commission announces annually, and compare the transaction behavior difference and efficiency of the transaction effect. To directly review the change in the transaction performance of companies whose affiliations have changed from large to medium business groups due to changes in the large-size, business-group designation criteria, we perform an additional panel analysis using a business-group affiliation change dummy variable (CHANGE). By analyzing business performance using business-group designations that are related to government regulations, we investigate the effectiveness of the current regulatory policies. The main empirical analysis is based on a model that applies an affiliation dummy variable (GROUP). Our analysis compares the transaction efficiency between subsidiaries for each business-group designation. When the variables included in the regression model are considered to be corporate characteristic variables that can interact as factors to determine the size and behavior of the transactions between subsidiaries, it is possible for the internal transaction ratios (SR, PR, SPR) to become endogenous explanatory variables. Hence, we use the results of the two-stage least squares estimation method to control the endogeneity problem. We use the average value of the internal transaction ratio of industry, the number of business group-affiliated subsidiaries, and industry correlation as the instrumental variables for estimating the internal transaction ratio variables. The results are as follows. First, according to the regression analysis of the effect of sale and purchase transactions between subsidiaries on firm performance, the transactions between subsidiaries of a large business group subject to regulations have a significant, positive impact on business performance, and thus support the efficient-transactions hypothesis. In contrast, we find a significant negative correlation between the intra-group transaction ratio and firm performance in unregulated, medium-sized business groups. This evidence is consistent with the tunneling and inefficiency hypothesis, which supposes that large Korean business groups have used transactions with affiliates to illegally transfer owner's wealth and management rights. Interestingly, we can confirm that the inefficiency of unfair internal transactions is observed in medium business groups rather than in large business groups. This is contrary to common belief and the findings of previous studies on Korean Chaebol firms. Second, according to the results of the analysis of the intra-group transactions effect before and after the point of regulation change based on using a business-group affiliation change dummy variable, the transactions between subsidiaries of a company for which the affiliation has changed from a large business group to an unregulated medium business group have a significant negative impact on firm performance. We interpret this result as follows. The change in affiliation of a regulated business group to a non-regulated business group has a significant negative impact on the effect of transactions between subsidiaries. This study contributes to the literature on business groups by comparing the firm performance between two groups, before and after regulation change. In summary, differences in the efficiency of transactions between subsidiaries are revealed based on their affiliation with a large business group and the change of affiliation to a medium business group. In other words, the efficiency of internal business transactions depends on whether the firm belongs to large business group or not, and the business group's regulation designations are very important. The findings of this study thus provide the following suggestion for government policies: rather than the ex-ante, uniform method for regulating transactions between subsidiaries based on the large business group designation system, institutional supplementation is needed in the form of an ex-post regulation method based on the expansion of the target range for regulation and judgment of appropriateness. If restriction is uniformly applied to all transactions between subsidiaries based on the large business group designation system, it may constrict even efficient business management activities. Hence, the appropriateness judgment criteria for evaluating the efficiencies of individual transactions between subsidiaries are needed, and rather than the method of regulating ex-ante internal transactions outright, ex-post regulation based on the actual content of individual internal transaction is desirable. It is necessary to adjust a range of regulated firms and to complement monitoring systems for unfair trade practices, based on ex-post evaluations.