The coupling of inductively coupled plasma mass spectrometry (ICP-MS) with spark ablation (SA) was applied for the semi-quantitative determination of impurities in different types of ferroalloys. The sparking operating conditions were optimized, and a restrictive path including a cyclone and a spray chamber was used to decrease the quantity of sparked material reaching the plasma, to prevent torch injector and sampling cone blockage. Samples were mixed in a 1:2 sample-to-graphite ratio and pressed into pellets for direct solid sampling by the spark. The advantage of the technique includes the benefit of easily determining some refractory elements that are very difficult to dissolve. The method was tested on three different types of ferroalloys: Fe-Ti, Fe-Nb, and Fe-V. A single standard ferroalloy sample was used as an external calibration sample to create a complete response curve for each type of ferroalloy base. Industrial ferroalloys, chemically characterized by using different analytical methods, were applied as calibration samples. The use of a calibration sample closely matching the ferroalloy material to be analyzed was found to be essential if accurate analyses were to be obtained. Certified Reference Materials of the three ferroalloys types tested—BAS (Bureau of Analyzed Samples) 579-1 (Fe-Nb alloy); BAM (Bundesanstalt für Materialprüfung) 589-1 (Fe-Ti alloy); and BAS 577-1 (Fe-V alloy)—were analyzed to evaluate the accuracy attainable in this panoramic mode. Most of the results, ranging from 0.05% to 5% m/m, were observed to be accurate to within 6–18.6% of the certified value, and the precision was better than 17.8% relative standard deviation. Determination limits based on 10 times the standard deviation of six replicates of a blank graphite pellet were on the μg g−1 level.