<p>Over the last sixty years there has been disagreement about the meaning of angels holding musical instruments in Trecento painting. Some regard them as simply symbols of heavenly music" others argue that payments are documented to instrumentalists 'when they made like (<i>fecerunt</i>) angels'. To resolve this issue I have used the quasi-complete catalogue of all Trecento paintings compiled by the late Howard Mayer Brown as a basic data set of images, and two inventories derived from Trecento literature, one of the names of musical instruments, the other of musical 'activities' and generic names of musical performance. From these three sources, I hoped to document musical instruments in pictures, payment records, contemporary chronicles, imaginative literature and music theory treatises.<p> <p>Chapter 1 reviews current literature on the topic" Chapter 2 reviews the production and function of objects on which the images were painted" Chapter 3 analyses the iconography of Brown's <i>Catalogus</i> and two iconic images, the Coronation of the Virgin and the Virgin and Child (sometimes called <i>Maesta</i>, a Tuscan version of the <i>Herogetria</i>). The fourth chapter examines evidence of music activity in Florence: <i>Musici</i>, the performer-composers" itinerant musicians or <i>giullari</i>" the professional musicians salaried by the government" amateur musicians like the ones described by Boccaccio in his <i>Decamerone</i>" clerics, all of whom had some musical training for the performance of the liturgy, some who cultivated the skill of improvised counterpoint and some who composed the music found in secular manuscripts" and finally, <i>Laudesi</i>, the confraternities who gathered regularly before an image of the Virgin Mary to sing her praises (<i>laude</i>). Chapter 5 examines in detail the contribution written music may make to the use of musical instruments in performance. A conclusion follows summarising the findings of this thesis and their significance for further research.</p> <p>After the examination of 368 paintings and 41,493 texts from a database of 48,985 records (3,266 from chronicles" 25,634 from payment records" 4,151 from literary works and 342 from musical treatises), no unequivocal evidence has been found for the use of musical instruments on the untexted voices in the surviving manuscripts. Yet, there is much circumstantial evidence: professional instrumentalists were paid to accompany the <i>laude</i>" three-voice works are written into vacant folios of manuscripts written at the same time as local and visiting <i>pifferi</i> were being paid by the Signoria. Untexted contratenors in works by Francesco degli organi (also known as Francesco Cieco or Francesco Landini) and Don Paolo Tenorista may have been intended for an instrument, and this may have been the third kind of music referred to by Filippo Villani (vocal music, instrumental music and a mixture of the two). Some of the composers are documented as instrumentalists: might not Jacopo da Bologna, Giovanni Mazzuoli and Francesco degli organi have played their own music? We have no evidence, one way or the other. Nor is there any evidence that untexted voices were vocalised. All circumstantial. On the other hand, there is no positive evidence to exclude instrumental performance of written music. The evidence for angelic musicians is more secure. Goro Dati remarks on processions of 'confraternities of laymen who come together . . .with the clothing of angels' (<i>con abito d'angioli</i>)" lauda singers were sometimes paid 'when they dressed like angels' (<i>quando si fecerunt angioli</i>)" and the confraternity of S. Zenobi at the Cathedral possessed 'six garments of angels . . . and six pairs of angel's wings' (<i>6 chamici da angioli . . . 6 ispalliere da angioli</i>).</p> <p>After a thorough examination of all of the paintings, all of the surviving published payment records and a generous selection of contemporary literature, no unequivocal evidence for instrumental participation in secular polyphony has been found but the circumstantial evidence supports it. There is strong evidence that the angel musicians performing before the iconic image of the Virgin and Child were images based on the lauda service: real people making real music.</p>