Though we have had extensive theoretical and empirical studies on diversification during the past decades, yet the impact of diversification on a firm’s financial performance remains unclear. Earlier, authors (like Arnould, 1969; Berry, 1971; Gort, 1962) tried to answer the fundamental question of ‘whether a firm should diversify or not’, but were unable to reach any consensus. Rumelt (1974) categorized diversification into related and unrelated and concluded that diversification in a related area is better than being undiversified. Even after the seminal work of Rumelt, empirical evidence on the impact of both types of diversification on a firm’s financial performance is still mixed (Berger & Ofek, 1995; Chen & Joseph Yu, 2012; Duin & Hansen, 1991; Palepu, 1985; Palich, Cardinal, & Miller, 2000). In this study, we make an attempt to answer the same fundamental question of ‘whether a firm should diversify or not’ by including three new aspects: first, we measure the impact of diversification (and its types) on the three aspects of a firm’s financial performance, that is, risk, return and risk-adjusted return; second, we measure this impact on lag 1 of diversification; and third, we use a newly developed approach, that is, correlation-based diversification measures (Nigam & Gupta, 2018b) to measure different types of diversification. Initially, our results indicated insignificant impact of diversification (and its types) on all firm performance measures. Later, we segregated related diversification (RD) into positive related diversification (PRD) and negative related diversification (NRD); then we measured the impact of each type of diversification separately and found that diversification is better than being undiversified only if it is into a negative related area. It is a new finding and may have some policy implications for the management while designing its diversification strategy.