Oscillator strength calculations for the transition elements are complicated by the occurrence of many strongly coupled terms from a single configuration. For example, there are nine ²F^o terms in the 3d⁴4p configuration of the vanadium sequence. Using the specific example of the transition 3d⁴4s ⁴H to 3d⁴4p ⁴G^o in Cr II, it is shown that calculations having poor energy separations for the four 3d⁴4p terms are very unreliable but, if corrected for energy spacing, a calculation using reasonable but rather simple wavefunctions is as good as one using considerably more elaborate forms.