Nicholas Southerland appeals his conviction of aggravated assault with a firearm. He contends that the trial court erred in not excluding the results of a neutron activation test after finding that the State had committed a discovery violation by not producing the test results until the last day of trial. We affirm.

On November 8, 1999, Ellis Wayne Brit informed police officers Strickland and Harris that he and Brenda Lee Lane had just been shot at twice by Southerland at a trailer park about a half a mile away. The officers proceeded to the trailer park with Brit, who identified Southerland as soon as they pulled into the driveway of the park. The officers yelled at Southerland to stop, but he ran. The officers very quickly found him crouched down underneath a trailer with something in his hand. The officers took Southerland into custody and retrieved a gun from underneath the trailer within inches of where Southerland had been. Officer Harris, who has had extensive experience with firearms during his seven years as a police officer and four years as a military firearms instructor, testified that he smelled freshly burnt gunpowder when he retrieved the gun, which told him that the gun had been recently fired. He also observed that two successive rounds of ammunition had been fired.

On the day of trial, May 22, 2000, the State informed the court that, although it had ordered a neutron activation test for the presence of gunpowder residue on Southerland’s hands by letter to the Florida Department of Law Enforcement dated February 3, 2000, the test had never been done. The State became aware of this fact when defense counsel requested the results of the test on May 16, 2000. The explanation was that FDLE did not perform the test because they thought that Southerland’s case was going to be removed to federal court. The trial court was able to schedule the taking of testimony such that the twenty-hour test could be completed before the trial ended.1 On the *156last day of trial the FDLE witness was able to make it to Tampa from Orlando with the test results and testify as the last witness in the case.

The State learned that the results of the neutron activation test were inconclusive before the FDLE witness arrived to testify on the last day of trial. The defense objected to the admission of the results on the ground that the State had committed a discovery violation by not having the results before trial. The court held a Richardson2 hearing and concluded that the State had committed a discovery violation but that it was not wilful. It therefore decided not to exclude the test results but to allow defense counsel to depose the witness before she testified.

The FDLE witness testified over defense counsel’s continuing objection that three particles of gunshot residue were swabbed off of Southerland’s hand. She further stated, however, that this amount is considered forensically insignificant because a positive result for the presence of gunshot residue requires the presence of four particles.3 Nevertheless, the State asserted during closing: “And folks I want you to also remember this, gun powder found on the defendant’s hands while not up to the four particle standard, it was three particles.”

We agree with the trial court’s handling of the discovery issue in this case. There was no evidence that the State wilfully prevented the running of the neutron activation test,4 and it certainly did not hold back the results of the test.5 Even if we found that the trial court abused its discretion in not excluding the neutron activation test results, there was no harm to Souther-land in light of the compelling evidence of his guilt. Specifically, he was apprehended moments after running from police officers who saw an object in his hand and then found a gun that had recently been fired twice within inches of where Souther-land was found.

Accordingly, we affirm Southerland’s conviction.

Specially concurring.

I concur in the decision because I agree that the evidence of guilt was overwhelming and, thus, the error was harmless. I also agree that the results of an inconclusive test would be irrelevant and thus inadmissible if not for the precedent from our supreme court. See Mills v. State, 476 So.2d 172 (Fla. 1985).

I write because I would hold that it was error to admit the test results when the State failed to obtain this evidence until the last day of trial and Southerland’s defense preparation was prejudiced by the State’s action (or inaction, as the case may be). Clearly the results of the “inconclusive” test were prejudicial as evidenced by *157the bolstering of the State’s case in final argument. I remain perplexed over why-prosecutors and trial judges would risk a mistrial or new trial because of questionable evidence when the remaining evidence was overwhelming as to the defendant’s guilt.