Ignition tests were carried out using three different laser systems and three different pyrotechnic compositions. Pyrotechnic materials investigated are: sulfur/charcoal/potassium nitrate based composition (gunpowder, GP), Shellac binder-based boron/potassium nitrate composition (SR 44) and acaroid resin binder based magnesium/potassium nitrate composition (SR 371C). The laser sources were the multimode output from an Ar-ion laser (lambda = 500 nm average), a high-power commercial diode laser (lambda = 784 nm) and a small laser diode operating at around the same wavelength but controlled by a customized electronic circuitry. Lasers operating in the visible wavelength range provided more reproducible and quicker ignition than the infrared output from the diode lasers. It was found that unconfined gunpowder exhibits more reproducible ignition for both the visible and the infrared wavelengths compared to the other two compositions. The composition based on magnesium, SR 371C appeared to be very sensitive to laser intensity variations and gave erratic and therefore, irreproducible ignition delay times. The threshold laser energies to initiate reproducible ignition for the different wavelengths were measured and ignition maps were constructed. From these maps, the required laser power density for any value of the ignition delay time, i.e. laser energy density was determined. Tests were also conducted on gunpowder samples, partially confined in a modified pyrogen igniter capsule and a small laser diode. The diode was operated in single pulse mode using a current surge, which was much higher than the recommended value for CW operation. This provided similar to 1 W pulses at the end of a I mm diameter fiber optic cable and caused reproducible ignition in the semi-confined pyrotechnic bed within the capsule. The threshold ignition energy under semi-confined conditions was found to be substantially less than that required in the unconfined environment under similar experimental conditions.